Audio Codec Transform Length Switching for Transient Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing frequency-domain audio codecs struggle to provide high-quality encoding for audio signals with prominent transients, such as rain or applause, at low bitrates, due to issues like time-smearing and increased data overhead, and switching between transform lengths is challenging without affecting backward compatibility.
Innovation Solution
A frequency-domain audio codec is enhanced to support additional transform lengths by interleaving frequency-domain coefficients and scale factors independently of signalization, allowing existing codecs to operate without faults while achieving better quality with minimal coding efficiency penalties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If long transform blocks are used for encoding, then spectral resolution is improved, but time-smearing and pre-echo artifacts increase for signals with prominent transients
Solution Approach 1:
The audio signal is divided into multiple transform blocks of different lengths (short, medium, long) that can be selectively applied to different time regions. This segmentation allows the encoder to use long blocks for stationary regions (achieving high spectral resolution) while using short blocks for transient regions (avoiding time-smearing), thus resolving the contradiction between spectral resolution and time-smearing artifacts.
2Object-generated harmful factors
If short transform blocks are used for encoding, then time-smearing is reduced, but spectral resolution decreases and data overhead increases
Solution Approach 1:
The transform block length is made dynamic rather than fixed, allowing the encoder to adaptively select between short, medium, and long blocks based on the local signal characteristics. This dynamic adjustment enables the system to use short blocks only when necessary (for transients) while using long blocks for the majority of stationary regions, thereby maintaining spectral resolution while reducing time-smearing artifacts.
3Adaptability or versatility
If multiple transform lengths are supported, then coding quality for various signal types is improved, but backward compatibility with existing decoders is compromised
Solution Approach 1:
The encoder is designed with multi-functionality to handle both legacy and enhanced decoding scenarios. The bitstream includes syntax elements that indicate transform block length information, allowing enhanced decoders to utilize multiple transform lengths for improved quality while ensuring that legacy decoders can still process the bitstream using default long-block processing, thus maintaining backward compatibility.
4Productivity
If transform block switching is implemented, then coding efficiency for transient signals is improved, but device complexity and implementation difficulty increase
Solution Approach 1:
The implementation leverages existing parameter structures in frequency-domain audio codecs by reusing scalefactor band definitions and quantization schemes across different transform block lengths. This parameter harmonization reduces implementation complexity while still enabling transform block switching to improve coding efficiency for transient signals.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A frequency-domain audio codec is provided with the ability to additionally support a certain transform length in a backward-compatible manner, by the following: the frequency-domain coefficients of a respective frame are transmitted in an interleaved manner irrespective of the signalization signaling for the frames as to which transform length actually applies, and additionally the frequency-domain coefficient extraction and the scale factor extraction operate independent from the signalization. By this measure, old-fashioned frequency-domain audio coders/decoders, insensitive for the signalization, would be able to nevertheless operate without faults and with reproducing a reasonable quality. Concurrently, frequency-domain audio coders/decoders able to support the additional transform length would offer even better quality despite the backward compatibility. As far as coding efficiency penalties due to the coding of the frequency domain coefficients in a manner transparent for older decoders are concerned, same are of comparatively minor nature due to the interleaving.